Not exactly what you're looking for but here's some stuff you can look
at....

1) Dispatching priority... set it according to how critical you view this
service... I've got ADSM set up with a dispatching priority just below TCPIP
(which has a high dispatching priority - about the same as VTAM)

2) Recovery log - Q LOG Format=Detailed gives a detailed report on your Log
Pool... check the Log Pool Pct. Wait: - if it isn't 0.00, increase
LOGPoolsize in your server options files.

3) Database - Q DB Format=Detailed gives a detailed report on you
Database... check the Cache Hit Pct and Cache Wait Pct... if the wait pct
isn't 0.00, consider increasing the BUFPoolsize.

4) DASD stuff - keep the Recovery Log and Database files on separate
physical volumes.  With 3990-3 controllers, caching isn't gonna buy you
anything (can even hurt).  3990-6 controllers with record level caching
might be a different story.

5) Background processing - try to run Expiration Processing, Migration, and
Reclamation at times that don't interfere with the client backups.
 Expiration processing is a CPU hog so I run it on weekends.  Migration is
run early in the morning and Reclamation is limited to 1 or 2 days a week
during the day.


Tim Pittson
pittson1 AT bwmail1.hcc DOT com
 ----------
From: owner-adsm-l
To: Multiple recipients of list ADSM-L
Subject: MVS ADSM Server Performance Tuning.
Date: Tuesday, February 13, 1996 6:44PM

Hi All,

I would appreciate hearing from anyone who has experience with tuning an
MVS/ESA ADSM
Server running on an 3090 cpu.  I'm aware that increasing the size of the
recovery log
(at least in version 1) can help as it alleviates ADSM from doing a lot of
I/O to the
database by increasing the odds of a clear page being available in the
recovery log.
I have not tried this yet on the problem system as I would like to see if
there is a
methodology to determining what is going on with the server so I understand
exactly what
I'm trying to fix.

Would running some of the server trace options help?  I currently have a
ADSM server
running on a larger cpu (than the 3090) that uses less than 1% of cpu and I
have an ADSM server running on the 3090 using 6% (idle, I think...) -
wondering if I can lessen the
disparity between the 2 (I realize the larger machine is much more
powerful...).

Thanks,
Dave.

 ---- message ------